Despite the development of oral hypoglycemic agents that boost insulin secretion or reduce insulin resistance, and of different forms of insulin with widely varying durations of action, and the advent of a markedly improved protocol for islet transplantation, the treatment of insulin-deficient diabetes remains unsatisfactory. Experimental gene therapy is being explored as a possible option for an improved treatment or potentially a "cure" for the disease. A promising gene therapy strategy that was recently developed by the Principal Investigator is that of induced islet neogenesis by the delivery of islet-specific transcription factor to the liver of diabetic mice. We used a helper-dependent adenovirus (HDAd) to deliver a transcription factor, NeuroD/Beta2 (ND/B2), and an islet growth factor, betacellulin (Btc), to the liver of streptozotocin diabetic mice. In response to the treatment, the mice formed new pancreatic islets in the liver and developed normal blood glucose and insulin levels during a glucose tolerance test (GTT). The treatment reversed the diabetes of these mice for over four months, the duration of the experiment. There was negligible toxicity observed for the treatment. In this R21 application we propose to explore the use of a similar strategy to treat diabetes in a non-human primate model. There are 2 specific aims: 1. to establish an insulin-deficient diabetes model in baboons using streptozotocin (STZ) treatment. 2. To use a balloon catheter-based method to deliver into the liver of STZ-induced diabetic baboons HDAd containing the islet-specific transcription Neurogenin 3 (Ngn3) and the islet growth factor Etc to induce islet neogenesis with the objective of reversing the diabetes of baboons. Preliminary experiments show that Ngn3 produces even better results than ND/?2. Other pilot experiments in baboons have allowed us to develop the balloon catheter-based protocol to deliver HDAds at a high dose directly to the liver without significant systemic dissemination or toxicity. This protocol would circumvent the substantial toxicity of HDAds that is observed when the vector is administered systemically. These aims are based on well documented preliminary experiments. If the treatment strategy ameliorates diabetes in baboons without significant toxic side effects, then it will be an important proof-of-principle experiment that will pave the way for possible clinical trials in the future. [unreadable] [unreadable]